A collection of machine learning projects in finance, retail, and customer analytics — covering predictive modelling, classification, forecasting, and interactive app deployment.
📖 Description
Machine learning allows me to go beyond describing the past — it helps me predict and shape the future of finance, business, and customer insights.
From predictive modelling to classification and clustering, these projects show how I apply machine learning to:
- Financial data
- Customer behaviour
- Forecasting challenges
Each project demonstrates both the technical side (data prep, feature engineering, algorithms) and the business value (insights, risk management, decision support).
🛠 Tools I use across these projects: Python (pandas, scikit-learn, XGBoost), SQL, Streamlit, Power BI, and Advanced Excel.
Want to see my methodology and thought process?
I have documented detailed notes on my workflow, challenges, and key learnings while developing these projects. 📚 Read My Machine Learning Notes — workflow, challenges, and key lessons documented step-by-step.
✨ Featured Projects
💎 Diamond Price Prediction: The 4C Model
A predictive modelling project applying regression and ensemble learning to diamond valuation.
- 🔍 Feature Engineering – PCA, polynomial transformations, and scaling for robust modelling
- 📊 Model Comparison – evaluating regression, ensemble, and boosting methods
- 🧮 High Accuracy Forecasting – achieved R² = 0.982 (98% predictive accuracy)
- 🛠️ Deployment – interactive Streamlit app for real-time diamond price prediction
Click to dive into the full portfolio and see how I connect Python, scikit-learn, and XGBoost to deliver an end-to-end machine learning pipeline with interactive deployment.
⚖️ SuperStore Sales Analysis & Profit Forecasting
A financial analytics project exploring sales drivers, discount optimisation, and profit forecasting.
- 🔍 Sales & Profit Insights – analysing patterns across products, regions, and discounts
- 📊 Impact Quantification – measuring how pricing and discounting affect profitability
- 🧮 Predictive Modelling – classification models achieving 85%+ accuracy on profit prediction
- 🛠️ BI Dashboards – Power BI visualisations for strategic decision-making
Click to dive into the full portfolio and see how I connect Python, scikit-learn, and Power BI to deliver actionable insights and forecasting for retail profitability.
⚡ Continuous Learning
- I continue to add new projects as I explore advanced techniques in anomaly detection, deep learning, and financial forecasting.
- Connect with me on LinkedIn or explore all repositories on GitHub.
© Teslim Adeyanju 2025. All Rights Reserved.